Social Media Sponsorship: Metrics for Finding the Right Content Creator-Sponsor Matches
76 Pages Posted: 5 Dec 2019 Last revised: 12 Oct 2020
Date Written: November 19, 2019
Social media video sponsorship, in which a sponsor forms a partnership with a content creator and sponsors video content, has become increasingly popular. Although the extant theoretical literature supports the importance of the congruency between the sponsor and sponsee in forming sponsor-sponsee matches, there has been limited empirical research on the actual construct of sponsor-sponsee congruency. Using rich data on sponsored and non-sponsored videos on Facebook, we build upon the associative link and complex network frameworks to introduce two marketing metrics, Content Similarity and Audience Closeness, which can help marketers to find effective creator-sponsor matches that generate high audience exposure. Content Similarity extends the Latent Dirichlet Allocation (LDA) model to measure video topic similarity between creators and sponsors. Audience Closeness measures how close two nodes (i.e., creator and sponsor) are in the network via Dijkstra’s algorithm. Content Similarity has positive effects on video views, and Audience Closeness has nonlinear and diminishing effects on video views. The interaction effects between the metrics reveal that the positive effect of Content Similarity on video views is stronger for sponsor-sponsee pairs with low Audience Closeness, and vice versa. The metrics can also be used in combination with other marketing strategies for synergistic interaction effects.
Keywords: Social Media Video Sponsorship; Content Creator; Sponsor; Sponsorship Relevance; Natural Language Processing; Latent Dirichlet Allocation; Machine Learning; Matrix Completion
JEL Classification: M31; M37
Suggested Citation: Suggested Citation